Whereas we have made significant progress in demonstrating the power of digital microfluidics in sequencing by synthesis in our current R21 proposal, some key experimental aspects of the technology as applied to achieving long reads need to be demonstrated. Thus, the overarching aim of Year 1 is to extend the read length of the droplet based sequencing-by-synthesis pyrosequencing reaction chemistry by incorporating the following improvements: 1) integrate solid-phase attachment of DNA compatible with repeated droplet washing and exposure to reagent droplets;2) demonstrate synthesis reaction chemistry in a droplet format immersed in a silicon oil medium;3) interfacing of the electrowetting chip with off-chip reagent supply sources to achieve uninterrupted 350 base pair reads. Our specific aims for Years 2 and 3 of this proposal are: 1) determine read length and throughput limitations using adaptive reagent delivery strategies with feedback control based on the detected light signal. Demonstrate that homopolymer regions can be sequenced through with no or negligible degradation of the subsequent sequence accuracy;2) extend the simulation capability to develop a physically based model to estimate the achievable accuracy and use the insights to increase the read length of droplet-based pyrosequencing;3) develop a strategy and architecture for parallel reactions and form estimates for the electrowetting chip area, CMOS photodetector array size, and electrode size needed to scale to our 10 year goal of 10,000 parallel reactions;4) demonstrate that electrowetting technology can be scaled to a picoliter droplet format by scaling system dimensions to achieve highly parallel reactions;5) experimentally determine read length limitations in droplet-based sequencing by synthesis, and implement software and signal processing strategies to improve read lengths and data quality and throughput, with a goal to demonstrate 1,000 to 10,000 base pair reads by Year 3. The research design is based on verifiable subtasks for each aim that are driven by group leaders. Two key inventions underlie our proposed modified pyrosequencing approach to obtain long reads: 1) decouple the sequencing and detection steps use feedback to add separately extra nucleotides at any DNA site where homopolymer regions are encountered and detected;2) utilize droplet-based electrowetting to handle the massively complicated fluid handling problem.